In recent years, a considerably concentrated grocery retail sector has brought to the fore the possible consequences of retailer buyer power on upstream firms. A retailer can, for example, delist a product to punish a particular supplier for not meeting its requests. It might as well use delisting as a threat, to get better terms of trade. How will consumers react to the delisting of a product? We develop a model of demand for multiple stores and products in which the key modeling feature is to explicitly account for the observed heterogeneity in consumers’ shopping patterns by including consumers’ transaction costs of shopping —real or perceived costs of dealing with a store. We use scanner data on grocery purchases by French households in 2005. We then estimate the parameters of the model and recover the distribution of shopping costs. Preliminary results show that the total shopping cost per store sourced is 4.68 e per store sourced on average. This cost has two components, namely, the mean fixed shopping cost, 2.95 e, and the mean total transport cost per trip, 1.73 e per trip. We show that consumers able to source three or more grocery stores have zero shopping costs, which rationalizes the low proportion of three-stop shoppers observed in our data. In the next iteration of the paper we will use the estimated model to simulate the delisting of one of the inside products from only one store out of three.
Acerca del expositor
Jorge Florez Acosta es economista de la Universidad de Antioquia. Obtuvo su doctorado en Economía en 2015 en la Escuela de Economía de Toulouse en Francia. En la actualidad se desempeña como profesor de la Facultad de Economía de la Universidad del Rosario. En el pasado se desempeñó como profesor del Departamento de Economía de la Universidad de Antioquia. Sus áreas de interés son la Organización Industrial, la Econometría Aplicada, la Política de la Competencia y la Teoría Microeconómica Aplicada.